HD1960MM1SV

Raymond Atchley Jet Pipe Servo Valve:

Fig MM1\ATCH-9&8

Fig MM1\ATCH-7

The Atchley valve made a big hit when it came on the market. In fact LTV used it on their first MM1 servo actuators, above right. However in time the Moog valve prevailed for Production.

The torque motor moves a small tube squirting oil into one of two receiver ports to move the valve spool, which in turn ports fluid to the controlled actuator. This and the Moog flapper valve were subject to contamination – Moog solved that by building a “dry” torque motor valve which separated electrical and oil.

These valves were driven by vacuum tube amplifiers and thus limited to 8 ma max valve current. At the time there was no way to drive a hydraulic control spool directly with electronics. Thus it was necessary to devise ways to amplify control with hydraulics. However this came at a penalty. There was a constant flow of hydraulic leakage, to achieve fast response the oil could not be shut off.

The Vacuum Tube driver left was often augmented by

Dither stage right to keep the valve active and not silt stuck with contaminates.

Transistorized Valve Driver AC to DC Power supply for Valve Drivers

Rack Mounted Servo Valve Controller

Vacuum Tubes were the state of the art at the start of MM 1. Autonetics was one leading the way applying transistors to missile controls – aircraft systems would follow much later.

We used a commercial HiFi Amplifier, much larger than the above power supply, to provide quality Square Wave Excitation which emulated 5K hz chopped battery DC.

Digital Electronics replaces Analog to drive 8ma servo valves

About 1970, having switched to experiments with digital electronics, I made a valve driver from one RCA 4016, a new CMOS integrated circuit, which replaced the above drivers. The story of how that came about is separately described. These company supported experiments I did eventually caused me to change fields to electronics, at the close of rocket engine systems work. The methods evolved were integrated into the MX missile electronics which permitted shaping servo-loop performance by digital computation rather than analog shaping. The digital electronic method retained the highly evolved attributes of hydraulic servo actuators – putting to rest what was once perceived as a need to build “digital servo actuators”. Driving analog servo valves with a single chip was the first breakthrough. More difficult was the conversion to digital of the highly reliable Linear Inductive Transformer position transducer output. These transducers, initially designed by Gary Collins an emulation of radio IF transformeres, swept the aerospace industry like wild fire and remain the primary method in current (2000 & on) avionics.

Vickers Flywheel Driven Hydraulic Pump

The first I knew of such a thing as Minuteman was in a conversation with Art Grear in the company cafeteria in Downey CA. I was operating the Extreme Temp Test facility and Art had been a lead systems engineer on Navaho. Art had recently been assigned to work on a proposal concept for Minuteman. Art knew I’d been an airplane mechanic and engineering officer on B-29’s as well as acquainted with various methods of setting up systems for test. He described the Minuteman as a set of three solid propellant boosters, and we proposed to control sets of tilting nozzles on the aft of each to perform attitude control – but there was no shaft power from which to drive a hydraulic pump. Did I have any ideas? I paused for a few minutes and then speaking as I thought I ran through a check list of methods used to start aircraft engines before there was any power. There were those that used shot gun shells to turn a starter, there were those like the P-40 and others where you cranked up a flywheel by hand or small electric motor and once up to speed you engaged the flywheel to start the engine. I said all that resulted in the B-29 method which was a hybrid, the armature of the starter motor also served as a flywheel. Art had not thought of using a flywheel before and asked if I had any other examples. I said every farm kid knew about cranking up the flywheel of a cream separator. In the days of single cylinder gas engines such as the wheat mill near us, they used large flywheels to fill in between power strokes. Also most Model-T drivers knew that if they had difficulty turning over the engine, they would jack up a back wheel and use it as a fly wheel. I jokingly said I guess you could spin one engine relative to the other and have plenty of power to run the system.

Thus I was not surprised when Art and Lou Purpura came out to the Test Lab with a flywheel system they had the shop build for them with a pump attached. For two months we ran many tests to determine how well such would work. In time we became convinced that, though it worked, it did not perform as required to meet our Minuteman needs. We reluctantly gave up what had seemed like a great idea. Dropping the idea was speeded up when they found they could obtain heavy duty light weight batteries – that a battery driven motor pump looked very good.

The concept had been presented by Art and Lou to TRW – technical advisors to the AF – and to Vickers with the idea of getting an experimental unit from them. It was not until several years later I learned that TRW and Vickers had pursued the idea by building hardware and making studies – after we had already dropped the idea.

Vickers experimental Flywheel system

Ron Frazinii made the final test runs after we obtained a better definition of what the duty cycle would be. Ron did a professional evaluation – we did not further work on the idea.

MOOG Static Load Error Washout and Majority Voting Servos

Fig MM2M-1 I don’t know what this 1962 missile is but it’s awesome.

I was so wrapped up in our world that I didn’t know this kind of thing was going on. It looks like an experimental version of an Intercontinental Liquid Rocket system – but too big to fit in a Silo. The G-38 Navaho missile had been cancelled a few years prior to this. So the only large liquid systems going that I knew of were the Atlas and the Titan and this is neither. This photo was without caption with a MOOG paper on “Static Load Error Washout” this is an obscure shorthand way of saying structural compliance (bending) had to be washed out in order to position the engine thrust for attitude control. Moog came up with their method – their servo actuators are circled above left.

Digital computer and associated developments had not come into play. Many were convinced that the only way to solve some of these problems was to cause the hardware to be smart – to both sense and control. At a later time sensors would instrument conditions and the digital computer work out commands to the servo.

Mathematical representation of the servo loop

Moog’s Voting Majority servo control applied to the F-111 fighter bomber

The beauty of the Minuteman -- was it’s refined simplicity.